Kinematics for 5-axis CoreXY printer

[tcpl]

Hi @dc42
I am in the process of building a 5-axis printer, with Core-XY linear kinematics with additional AB rotary axes mounted on the gantry. Could you reserve a kinematics type number for this purpose?
Thanks,
Blaise

[dc42]

@tcpl I can reserve a kinematics number, but do you actually need new kinematics? Unless you want the positioning of the CoreXY motors to take account of the position of the rotary axes, or vie versa, you don't need one because you can already add additional axes to a CoreXY configuration.

[tcpl]

@dc42 I need to take that into account, as I want GCODE to dictate the position of nozzle and it's rotation, not axes individually. This makes linear axes affected by target rotation, and I don't see how I could make that work with what is available now

[dc42]

@tcpl thanks. In firmware 3.7 we'll be shifting to a new mechanism for defining kinematics, so what we'll need is a unique short name of the kinematics to select it in the GCode command. Would you like to propose one? Syntax will probably be:

M669 K"your_kinematics_name"

where as usual case of letters is not significant, and underscores and hyphens are ignored when matching.

[tcpl]

@dc42 The best name I (or my team at my students' association) could come up with (and I have been working on this project for some time) is ABCoreXY. If someone has better name I would be more than willing to use it, but if not it is (semi)appropriate

[o_lampe]

@dc42 Over the years we've seen many experimental printer kinematics (eg. tripteron) or even worse: mixed kinematics like this.
IMHO there should be an option to define a new kinematic in RRF3.7 which includes naming it.

Maybe you can split the kinematic definiton into small sections?

• One for the gantry (delta, cartesian, H-bot, coreXY, Markforge, etc)
• one for the tool (static, AB rotary, etc.)
• one for the bed (static, Stewart, AB rotary, etc)

With that, the user can hopefully cobble together a kinematic model that fit's any flavour.